Patient-Specific QA for HDR Brachytherapy Skin Treatments
G Palaniswaamy*, B Massingill, N Deb, S Mutyala, D Rangaraj, Scott & White Hospital, Temple, TXSU-E-T-214 Sunday 3:00PM - 6:00PM Room: Exhibit Hall
Purpose: HDR treatments require high precision and accuracy but with current techniques and processes, there is higher probability for error. So we propose performing patient-specific QA for HDR skin treatment plans before the first treatment to eliminate the possibility of errors on first fraction that might lead to partial treatments or treatment delays.
Methods: A CT simulation is performed for skin patients before a week of their first fraction. The applicator is built at the time of simulation using HAM surface applicator. The applicator is then retained at the clinic to check the catheter orientation during treatment plan. There are several processes that are indirectly connected which increase the chance for errors that might propagate from the treatment plan to the patient's treatment. In the worst-case scenario, it might lead to partial treatment being delivered to the patient. These errors can be avoided by performing patient-specific QA before the first treatment which involves the following steps.
1. Catheter length measurements made independently by two personnel and verification at the treatment console.
2. Performing an autoradiograph with a film affixed at the outside edge of the applicator's target area.
Results: During 5 month period of treating 36 skin patients without patient-specific QA, we had 4/36 (>10%) instances of partial treatments and delays due to incorrect catheter lengths. After implementing QA, there were 0/12 partial treatments and we also detected two instances of dwell positions activated outside target.
Conclusion: By establishing the proposed QA technique, we were able to verify and eliminate the following errors:
1. Incorrect catheter lengths specified in the treatment planning system.
2. Active dwell positions in an optimized treatment plan that are outside the target area.
3. Unintended source obstruction errors during the first fraction.
4. Possibility of delivering partial fraction due to the above specified errors.